I remember when Clam signed up to the forums in 2008 and right away posted his first level pack. If I recall correctly I managed to solve all the later levels, but I was really stuck on the first level of his pack (and I think I wasn't the only one). Considering Clam was completely new, there was some speculation whether there was just a mistake in the pack and the level was actually impossible. Of course there wasn't and we were just blind, having to see a clever trick and using it twice. Which is what to me is a characteristic feature of Clam's levels. Many authors, me included, stumble across some trick (or back in the early days glitch) and then often build a level around it, trying to disguise it. Clam was a bit different in that regard that often in his levels some kind of trick or move would be used multiple times, or variants thereof, without too much other fluff around it. Sometimes you'd have to figure the key idea and apply it, sometimes it would look like the main idea is clear, but there are subtleties or variations that you'd have to figure out that make up for the main challenge. Either way, I felt that solution-wise, his levels would often have a theme to them, making them feel more unique and memorable. Interestingly, I don't necessarily associate small levels with Clam, and the ones that spring to mind first are actually not that small.


And to add a bit of trivia to the Tame 13 solution:

At some point Clam posted the following challenge (maybe in the context of which skills do you absolutely need to solve each level): Solve Tame 13 with just 1 builder. I looked at it and was intrigued because I had tried it before but given up. The only way I saw this as being possible was to somehow make a way up one of the pillars. I put in some thought and work and in the end I did manage, saving 100% to boot, which I posted about in the thread, assuming I had found Clam's solution probably with just a minor tweak or so to save those last one or two lemmings. Now Clam was intrigued, because his solution definitely couldn't be tweaked to get 100%. It never came to my mind that you can send up 20 climbers and use the nuke percentage glitch.

Yeah, the claw is terrain, it's a mistake on the page.

Long time ago I wrote a tool to rip all the L2 graphics, and you can see what is what (see attached). The claw doesn't have any animation (and doesn't appear in the objects list), but appears as terrain.

Considering the old outdated maps that we sometimes choose to play, I feel like some randomly generated maps wouldn't be that much worse than those for the occasional round. NaOH was working on randomized maps at some point, at least a basic version of that shouldn't be too hard.

As for organizing the actual maps, if more custom content comes in (currently there's a lull, but who knows) it would probably better to come up with a scalable solution. One such thing could be the server collecting users' ratings of maps after a games, and/or statistics of how often a maps gets played, and the offer to browse maps by popularity or select a random map proportional to popularity. Of course that'd need quite a bit of extra infrastructure.

Tbh I tend to be able to find quite quickly the maps I'm looking for, so it doesn't seem like a big issue to me. And then there's also the search functionality, right?

That's exactly as intended, great job! Thanks for the comment!

There's a decent chance I might actually make it for this Sunday

EDIT: Got a lot of stuff going on, so don't be surprised in case I don't show up tomorrow after all.

Whenever I visited Simon, this one came up and we were mystified and pondering what the name could be about.

I didn't notice the skill pattern (always 1 or 10 of a skill), this is a great find!
Good work putting everything together, how did you come up with the idea?

Great mystery solved, case closed. So no obscure lore behind this one, after all.

Interesting solution, though it looks super precise.

I don't have time to work on a fix within the next week at the very least, but Icho's suggestions sounds good. I remember contemplating how far to go with that considering the ease of execution/backroute safety trade-off.
If someone can go ahead with a fix that'd be great, especially if the execution of the intended solution doesn't get too tough as a result.